The Global Potential of Bioenergy onAbandoned Agriculture LandsJ . E L L I O T T C A M P B E L L , * , † , ‡

D A V I D B . L O B E L L , § R O B E R T C . G E N O V A , †

A N D C H R I S T O P H E R B . F I E L D †

Department of Global Ecology, Carnegie Institution ofWashington, Stanford, California 94305, Department ofBiological Sciences, Stanford University, Stanford, California94305, and Program on Food Security and the Environment,Stanford University, Stanford, California 94305

Received January 7, 2008. Revised manuscript receivedApril 7, 2008. Accepted May 22, 2008.

Converting forest lands into bioenergy agriculture couldaccelerate climate change by emitting carbon stored in forests,while converting food agriculture lands into bioenergyagriculture could threaten food security. Both problems arepotentially avoided by using abandoned agriculture lands forbioenergy agriculture. Here we show the global potential forbioenergy on abandoned agriculture lands to be less than8% of current primary energy demand, based on historical landuse data, satellite-derived land cover data, and globalecosystem modeling. The estimated global area of abandonedagriculture is 385-472 million hectares, or 66-110% of theareas reported in previous preliminary assessments. The area-weighted mean production of above-ground biomass is 4.3tons ha-1 y-1, in contrast to estimates of up to 10 tons ha-1

y-1 in previous assessments. The energy content of potentialbiomass grown on 100% of abandoned agriculture lands is lessthan 10% of primary energy demand for most nations inNorth America, Europe, and Asia, but it represents many timesthe energy demand in some African nations where grasslandsare relatively productive and current energy demand is low.

IntroductionAgricultural land dedicated to bioenergy crops is expandingrapidly (1), with multiple consequences for global climate,ecosystems, and food security (2). These consequences areclosely tied to the land that is used for bioenergy crops. Usingfood agriculture lands for bioenergy agriculture couldincrease the cost of the food commodities that are criticalto the diets of food-insecure people worldwide (3–5). Clearingforest land for new bioenergy crops could result in CO2

emissions from terrestrial carbon pools that are much greaterthan any greenhouse gas benefits provided by biofuels (6–11).Raising bioenergy crops on agriculturally degraded andabandoned lands is emerging as a sustainable approach tobioenergy that provides environmental benefits and climatechange mitigation without creating food-fuel competitionfor land or releasing the carbon stored in forests (12, 13).These lands have been defined as areas that have been

abandoned to crop and pasture due to the relocation ofagriculture and due to degradation from intensive use (14, 15).Growing conventional crops on these lands as a bioenergyfeedstock could increase rates of erosion and polluted runoff(16), while field studies suggest that growing low-input,perennial grasses as a feedstock would likely reduce suchimpacts (13, 17).

Assessing the global potential of bioenergy productionfrom agriculturally degraded and abandoned lands is chal-lenging because of the high uncertainty associated with thespatial extent of these lands and the potential plant produc-tion on these lands. Existing assessments have provided onlyrough global estimates of the potential bioenergy from theselands, using estimates of the global area that are ultimatelybased on expert opinion (18–20), the assumption of a homo-geneous spatial distribution, and spatially invariant or highlyaggregated estimates of plant production. Such estimatesrange from 430 to 580 Mha of abandoned agriculture land,with mean global plant production on these lands of 1-10tons aboveground biomass (AGB) ha-1 yr-1, meeting 2-23%of current global primary energy demand (13, 21, 22). Inprevious work, we developed a new method for estimatingthe abandoned areas that relies on historical land use dataand found that the potential biomass production has anenergy content of ∼5% of primary energy demand (23).Application of our data-driven approach at the national levelcould help inform policy makers of the potential scale of thisbioenergy resource (24).

Here we present a new global, spatially explicit estimate ofabandoned agriculture and the associated plant production onthese lands, using historical land use data, satellite-derived landcover, and global ecosystem modeling. We considered aban-doned agriculture as land that was previously used for crop orpasture but has since been abandoned (and has not beenconverted to forest or urban areas). The spatially resolvedbiomassproductionisthencombinedwithnational-levelenergyuse data to determine the potential contributions of bioenergy.

MethodsWe estimated abandoned agriculture areas using the histori-cal land use data from the History Database of the GlobalEnvironment 3.0 (HYDE, 5 min resolution) (25, 26). The HYDEgridded maps provide the fractional area of crop and thefractional area of pasture within each grid cell for each decadebetween 1700 and 2000. Abandoned areas were determinedfrom each map grid cell that had decreasing agriculture areasover time. Shifting agriculture, which also contributes toabandoned agriculture lands (15), is not included in thesemaps. As a check on this HYDE-based analysis, we alsoconsidered areas of abandoned crop from the Center forSustainability and the Global Environment (SAGE) land usedatabase (5 min resolution) (27, 28).

While the HYDE database provides the fractional areas ofboth crop and pasture, it does not provide the underlyingland use transitions (15). For example, if a grid cell shows adecrease in the fractional area of pasture and an increase incrop then it is not explicit whether these changes were dueto a land use transition from pasture to crop or from pastureto some other land cover within the grid cell. Understandingthis transition is important because transitions betweenpasture and crop should not be counted toward the area ofabandoned agriculture. We used two alternative approachesto estimate these land use transitions that provide a low andhigh estimate of total abandoned agriculture for each gridcell. In the more restrictive approach, providing the low

* Corresponding author phone: (650) 462-1047; fax: (650) 462-5968; e-mail: campbell@stanford.edu.

† Department of Global Ecology, Carnegie Institution ofWashington.

‡ Department of Biological Sciences, Stanford University.§ Program on Food Security and the Environment, Stanford

University.

Environ. Sci. Technol. XXXX, xxx, 000–000

10.1021/es800052w CCC: $40.75 XXXX American Chemical Society VOL. xxx, NO. xx, XXXX / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 APublished on Web 06/25/2008

estimate, we considered any simultaneous increase in cropand decrease in pasture to be a transition from crop topasture, and vice-versa. In the second approach, providingthe high estimate, we calculated abandoned crop area foreach grid cell as the difference between the maximumfractional crop area ever attained in a grid cell and thefractional crop area in the year 2000 (when this differencewas positive). We followed the same approach for calculatingabandoned pasture area. We then estimated the abandonedagriculture for each grid cell to be equal to the sum of theabandoned pasture and abandoned crop areas, but no greaterthan the difference of the total grid cell land area and thecurrent agriculture area. This is a more self-consistentdefinition of land use transitions than in our earlier study(23), which used MODIS land cover to constrain the land usetransitions between pasture and crop.

To exclude areas where abandoned agricultural lands havetransitioned to forest or urban areas, we used a MODISsatellite map of the current forest and urban land cover (3min resolution, MODIS/Terra Land Cover types MOD12C1)(29). Combining HYDE and MODIS data within this spatialanalysis can introduce errors due to the different assumptionsand input data used by the two data sets. Furthermore, theHYDE data provide subgrid cell information (fractional areas)while the MODIS data provide a single classification for theentire grid cell. We discuss the uncertainties in the followingsection. As a point of comparison, we also estimate theabandoned land area that has transitioned to forest or urbanareas based on the International Geosphere BiosphereProgramme (IGBP) DIScover data set (30, 31).

The biomass production at a specific location will dependon multiple factors including the crop type, management,climate, and soils. Natural production provides an upperestimateofpotentialagricultureproductionbecauseagriculturalharvest statistics, at a global scale, have been found to be about65% of natural production (32). To provide an upper envelopeestimate on the production potential, we used simulated naturalplant production from the Carnegie-Ames-Stanford Approach(CASA) ecosystem model (1 degree resolution) (33). Estimatingbiomass yield with the natural production model on a sitereflects local constraints from climate and soil types. Thesimulations (33) were driven by climate data (34), surfaceinsolation (35), soil texture (36), land cover (37), and thenormalized difference vegetation index (NDVI) (33). Usingsimulated natural production allows for the possibility that thetotal plant production from biomass agriculture may besignificantly higher than that for current agriculture, at the globalscale. The NDVI input may be sensitive to degradation whichmay decrease yields (13). We do not account for irrigation orvery high fertilizer inputs, which could increase yields (38–40).Above-ground production rates may also overestimate har-vestable biomass as some plant material must be left on theland to prevent further soil degradation (41).

Results and DiscussionBased on the HYDE historical land use data we found thatbetween the years 1700 and 2000, 269 Mha of crop lands werepermanently converted to land uses other than cropping (Figure1B), while 479 Mha of pasture lands were converted to landuses other than pasture (Figure 1C), at some point in the last300 years. This HYDE-based abandoned crop area is somewhathigher than the 210 Mha of abandoned crop area from theSAGE crop data (Figure 1A) (28). The abandoned crop areasfrom HYDE and SAGE data had the highest concentrations overthe Eastern United States, as a result of the relocation of croplandfrom the Eastern to the Midwestern region of North America.The most extensive area of abandoned pasture was over theMidwestern region of North America, where HYDE data indicatethat cropland has replaced pasture land. Australia, wherepasture areas peaked in the mid-1970s and have since steadilydeclined, also had high levels of pasture abandonment in thedata. We found that 99% of the land abandonment occurredduring the past 100 years.

Our low and high approaches to estimating the globalarea of abandoned agriculture (crop and pasture) yield totalareas of 474 and 579 Mha, respectively. These estimatesexclude abandoned agriculture areas arising from theconversion of crop to pasture or pasture to crop. Theseestimates do not exclude abandoned areas arising fromagriculture to forest or agriculture to urban transitions.Overlaying the MODIS land cover data, we identified thecurrent land cover classification of each grid cell containingabandoned agriculture areas. Excluding pixels classified asforest or urban areas yielded estimates of 385 and 472 Mhafor the low and high approaches, respectively (Figure 1D).

FIGURE 1. Global land areas of crop and pasture abandonment.(A and B) Crop lands that have been permanently converted toland uses other than cropping based on SAGE data andHYDEdata. (C) Pasturelands that have been permanently converted toland uses other than pasture based on HYDE data. (D) Totalabandoned agriculture based on HYDE data excluding areasresulting from land use transitions of crop to pasture, pastureto crop, agriculture to forest, and agriculture to urban (averageof higher and lower estimate).

B 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. xxx, NO. xx, XXXX

Our area estimates are 66-110% of the areas assumed inprevious preliminary assessments (13, 22, 42). Excludingadditional areas where abandoned agriculture has transi-tioned to other natural ecosystems types may also beimportant for maintaining species habitat and preservingnatural carbon pools.

Our spatial analysis, relying on both HYDE and MODIS data,introduces additional uncertainty due to the different methodsused in the creation of these data. Specifically, historical HYDEmaps were, in part, spatially distributed by human populationat the subadministrative level (25, 26). Since the subadminis-trative level distribution may be biased toward urban areas,our exclusion of MODIS urban areas may overcorrect ourabandoned area estimate. However, the abandoned agricultureareas were reduced by only 3% when using the MODIS data toexclude urban areas. Our application of the MODIS forest mapappears to have correctly excluded forest regrowth in the easternUnited States (Figure 1D) where abandoned agriculture hastransitioned to secondary forests (43). Alternatively, using theIBGP land cover data to exclude forest and urban areas resultedin a remaining abandoned agricultural area of 388-480 Mha.The IGBP area estimate is somewhat higher than the MODIS-based estimate of 385-472 Mha.

Using the global distribution of potential plant production(Figure 2A), we found that the abandoned agriculture landscould produce between 1.6 and 2.1 billion tons of AGB peryear for the low and high area estimates, respectively.Potential production rates on abandoned lands are highestin regions of tropical grasslands, ranging from 7 to 20 tonsAGB ha-1 y-1. Globally, the area-weighted average of theproduction rates on abandoned lands was 4.3 tons AGB ha-1

y-1. This is somewhat less than the global average of 4.9 tons

AGB ha-1 y-1 extrapolated from plot experiments (13) andin the middle of the range of 1-10 tons AGB ha-1 y-1 assumedin a preliminary modeling study (22).

The energy content of 1.6-2.1 billion tons of dry biomassis 32-41 EJ or 7-8% of primary energy demand. We assumeda high value for energy content of 20 kJ g-1

, although this

value will depend on multiple factors such as plant type andthe timing of harvest (44). At the national scale, the bioenergypotential was largest in the United States, Brazil, and Australia,where the available areas were the most extensive (Figure2B). The national bioenergy potential was less than 10% ofprimary energy demand for most countries in North America,Europe, and Asia while it represents many times the currentenergy demand in some African nations where grasslandsare relatively productive and current fossil fuel demand islow (Figure 2C). Converting the bioenergy crops to liquidfuels would cut the net energy to half this amount and couldresult in either a net greenhouse gas source or sink dependingon the types of agriculture and biorefineries used (45–48).

Overall, the potential bioenergy from abandoned agri-culture lands is a small but meaningful fraction of globalprimary energy consumption. Regionally, it can be moreimportant. The global potential could be increased withadditional land areas or through fundamental advances inbiomass agriculture. One possible source of additional landsis marginal agricultural lands that have limited potential forfood production. Land areas degraded from wood harvesting(49), rather than agriculture, are another possible source ofland. Forest lands could also produce net greenhouse gassinks for highly productive sites under sustainable harvestpractices (50). Consideration of these additional areas willrequire careful study of the competing uses, including, butnot limited to, food production, carbon storage in forests,and habitat conservation. Increasing yields of biomass cropsabove natural yields is another challenging path towardincreasing the potential of bioenergy. Based on abandonedlands and current technologies, however, expanding bioen-ergy crops to offset more than a small fraction of globalprimary energy consumption will present major challengesand difficult trade-offs.

Literature Cited(1) USDA. USDA Agricultural Projections to 2016; OCE-2007-1; U.S.

Department of Agriculture: Washington, DC, 2007; p 47.

FIGURE 2. Biomass production potential on abandoned agriculture lands. (A) Natural above-ground production of biomass on alllands determined from the CASA model, assuming 50% of the biomass is above-ground and the ratio (by mass) of biomass to carbonis 2.2. (B) Potential above-ground production of biomass on abandoned agriculture lands at the country level. (C) Ratio of the energycontent of the biomass on abandoned agriculture lands relative to the current primary energy demand at the country level. Theenergy content of biomass is assumed to be 20 kJ g-1.

VOL. xxx, NO. xx, XXXX / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 C

(2) Jordan, N.; Boody, G.; Broussard, W.; Glover, J. D.; Keeney, D.;McCown, B. H.; McIsaac, G.; Muller, M.; Murray, H.; Neal, J.;Pansing, C.; Turner, R. E.; Warner, K.; Wyse, D. Environment -Sustainable development of the agricultural bio-economy.Science 2007, 316 (5831), 1570–1571.

(3) Johansson, D. J. A.; Azar, C. A scenario based analysis of landcompetition between food and bioenergy production in theUS. Climatic Change 2007, 82 (3-4), 267–291.

(4) Leemans, R.; Eickhout, B.; Strengers, B.; Bouwman, L.; Schaeffer,M. The consequences of uncertainties in land use, climate andvegetation responses on the terrestrial carbon. Sci. China Ser.,C 2002, 45, 126–141.

(5) Naylor, R.; Liska, A.; Burke, M.; Cassman, K.; Falcon, W.; Gaskill,J.; Rozelle, S. Ripple effects of crop-based biofuels on globalfood security and the environment. Environment 2007, 49 (9),30–43.

(6) Fargione, J.; Hill, J.; Tilman, D.; Polasky, S.; Hawthorne, P. LandClearing and the Biofuel Carbon Debt. Science 2008, 219 (1235),1235–1238.

(7) Harmon, M. E.; Ferrell, W. K.; Franklin, J. F. Effects on carbonstorage of conversion of old-growth forests to young forests.Science 1990, 247 (4943), 699–702.

(8) Hooijer, A.; Silvius, M.; Wosten, H.; Page, S. PEAT-CO2,Assessment of CO2 Emissions from Drained Peatlands in SE Asia;Delft Hydraulics: Delft, The Netherlands, 2006.

(9) Houghton, R. A. The annual net flux of carbon to the atmospherefrom changes in land use 1850-1990. Tellus Ser. B Chem. Phys.Meteorol. 1999, 51 (2), 298–313.

(10) Righelato, R.; Spracklen, D. V. Environment: carbon mitigationby biofuels or by saving and restoring forests. Science 2007, 317(5840), 902.

(11) Searchinger, T.; Heimlich, R.; Houghton, R. A.; Dong, F.; Elobeid,A.; Fabiosa, J.; Tokgoz, S.; Hayes, D.; Yu, T.-H. Use of U.S.Croplands for Biofuels Increases Greenhouse Gases ThroughEmissions from Land-Use Change. Science 2008, 319 (5867),1238–1240.

(12) Hill, J.; Nelson, E.; Tilman, D.; Polasky, S.; Tiffany, D. Environ-mental, economic, and energetic costs and benefits of biodieseland ethanol biofuels. Proc. Natl. Acad. Sci. U. S. A. 2006, 103(30), 11206–11210.

(13) Tilman, D.; Hill, J.; Lehman, C. Carbon-negative biofuels fromlow-input high-diversity grassland biomass. Science 2006, 314(5805), 1598–1600.

(14) Daily, G. C. Restoring value to the worlds degraded lands. Science1995, 269 (5222), 350–354.

(15) Hurtt, G. C.; Frolking, S.; Fearon, M. G.; Moore, B.; Shevliakova,E.; Malyshev, S.; Pacala, S. W.; Houghton, R. A. The underpin-nings of land-use history: three centuries of global gridded land-use transitions, wood-harvest activity, and resulting secondarylands. Global Change Biol. 2006, 12 (7), 1208–1229.

(16) NRC. Water Implications of Biofuels Production in the UnitedStates; Committee on Water Implications of Biofuels Productionin the United States, National Research Council: Washington,DC, 2007; p 86.

(17) Tilman, D.; Reich, P. B.; Knops, J. M. H. Biodiversity andecosystem stability in a decade-long grassland experiment.Nature 2006, 441, 629–632.

(18) Grainger, A. Estimating areas of degraded tropical lands requiringreplenishment of forest cover. Int. Tree Crops J. 1988, 5, 31–61.

(19) Houghton, R. A. The future-role of tropical forests in affectingthe carbon-dioxide concentration of the atmosphere. Ambio1990, 19 (4), 204–210.

(20) Oldeman, L. R.; Hakkeling, R. T. A.; Sombroek, W. G. WorldMap of the Status of Human-Induced Soil Degradation. AnExplanatory Note (Global Assessment of Soil Degradation GLA-SOD); ISRIC: Wageningen, 1990.

(21) Berndes, G.; Hoogwijk, M.; van den Broek, R. The contributionof biomass in the future global energy supply: a review of 17studies. Biomass Bioenergy 2003, 25 (1), 1–28.

(22) Hoogwijk, M.; Faaija, A.; van den Broek, R.; Berndes, G.; Gielen,D.; Turkenburg, W. Exploration of the ranges of the globalpotential of biomass for energy. Biomass Bioenergy 2003, 25 (2),119–133.

(23) Field, C. B.; Campbell, J. E.; Lobell, D. B. Biomass energy: Thescale of the potential resource. Trends Ecol. Evolut. 2008, 23 (2),65–72.

(24) IEA. Biofuels for Transport: An International Perspective; In-ternational Energy Agency: Paris, 2004.

(25) Goldewijk, K.; Bouwman, A. F.; van Drecht, G. Mappingcontemporary global cropland and grassland distributions ona 5 by 5 minute resolution. J. Land Use Sci. 2007, 2 (3), 167–190.

(26) Goldewijk, K. K. Estimating global land use change over thepast 300 years: The HYDE Database. Global Biogeochem. Cycl.2001, 15 (2), 417–433.

(27) Goldewijk, K. K.; Ramankutty, N. Land cover change over thelast three centuries due to human activities: The availability ofnew global data sets. GeoJournal 2004, 61, 335–344.

(28) Ramankutty, N.; Foley, J. A. Estimating historical changes inglobal land cover: Croplands from 1700 to 1992. GlobalBiogeochem. Cycl. 1999, 13 (4), 997–1027.

(29) Friedl, M. A.; McIver, D. K.; Hodges, J. C. F.; Zhang, X. Y.;Muchoney, D.; Strahler, A. H.; Woodcock, C. E.; Gopal, S.;Schneider, A.; Cooper, A.; Baccini, A.; Gao, F.; Schaaf, C. Globalland cover mapping from MODIS: algorithms and early results.Remote Sens. Environ. 2002, 83 (1-2), 287–302.

(30) Belward, A. S.; Estes, J. E.; Kline, K. D. The IGBP-DIS 1-KmLand-Cover Data Set DISCover: A Project Overview. Photogram.Eng. Remote Sens. 1999, 65 (9), 1013–1020.

(31) Loveland, T. R.; Reed, B. C.; Brown, J. F.; Ohlen, D. O.; Zhu, J.;Yang, L.; Merchant, J. W. Development of a Global Land CoverCharacteristics Database and IGBP DISCover from 1-km AVHRRData. Int. J. Remote Sen. 2000, 21 (6/7), 1303–1330.

(32) Haberl, H.; Erb, K. H.; Krausmann, F.; Gaube, V.; Bondeau, A.;Plutzar, C.; Gingrich, S.; Lucht, W.; Fischer-Kowalski, M.Quantifying and mapping the human appropriation of netprimary production in earth’s terrestrial ecosystems. Proc. Natl.Acad. Sci. U. S. A. 2007, 104 (31), 12942–12947.

(33) Randerson, J. T.; Thompson, M. V.; Conway, T. J.; Fung, I. Y.;Field, C. B. The contribution of terrestrial sources and sinks totrends in the seasonal cycle of atmospheric carbon dioxide.Global Biogeochem. Cycl. 1997, 11, 535–560.

(34) Shea, D. J. Climatological Atlas: 1950-1979; National Centerfor Atmospheric Research: Boulder, CO, 1986.

(35) Bishop, J. K. B.; Rossow, W. B. Spatial and temporal variabilityof global surface solar irradiance. J. Geophys. Res. 1991, 96 (C9),16839–16858.

(36) Zobler, L. A. World Soil File for Global Climate Modeling;Technical Memo 87802; NASA, 1986; p 32.

(37) DeFries, R. S.; Townshend, J. R. G. NDVI-derived land coverclassifications at a global scale. Int. J. Remote Sens. 1994, 15(17), 3567–3586.

(38) Lee, D. K.; Boe, A. Biomass production of switchgrass in centralSouth Dakota. Crop Sci. 2005, 45 (6), 2583–2590.

(39) Lee, D. K.; Owens, V. N.; Doolittle, J. J. Switchgrass and soilcarbon sequestration response to ammonium nitrate, manure,and harvest frequency on conservation reserve program land.Agron. J. 2007, 99 (2), 462–468.

(40) Schmer, M. R.; Vogel, K. P.; Mitchell, R. B.; Perrin, R. K. Netenergy of cellulosic ethanol from switchgrass. Proc. Natl. Acad.Sci. U. S. A. 2008, 105 (2), 464–469.

(41) Lal, R. Crop residues as soil amendments and feedstock forbioethanol production. Waste Manage. 2008, 28 (4), 747–758.

(42) Hall, D. O.; Rosillo-Calle, F.; Woods, J., Biomass for energy: supplyprospects. In Renewable Energy - Sources for Fuels and Electricity;Johansson, T. B., Kelly, H., Reddy, A. K. N., Williams, R. H., Eds.;Island: Washington, DC, 1993.

(43) Albani, M.; Medvigy, D.; Hurtt, G. C.; Moorcroft, P. R. Thecontributions of land-use change, CO2 fertilization, and climatevariability to the Eastern US carbon sink. Global Change Biol.2006, 12 (12), 2370–2390.

(44) Domalski, E. S.; Jobe, T. L.; Milne, T. A., Jr. ThermodynamicData for Biomass Conversion and Waste Incineration; SolarEnergy Research Institute: Golden, CO, 1986.

(45) Adler, P. R.; Del Grosso, S. J.; Parton, W. J. Life-cycle assessmentof net greenhouse-gas flux for bioenergy cropping systems. Ecol.Appl. 2007, 17 (3), 675–691.

(46) Farrell, A. E.; Plevin, R. J.; Turner, B. T.; Jones, A. D.; O’Hare, M.;Kammen, D. M. Ethanol can contribute to energy and envi-ronmental goals. Science 2006, 311 (5760), 506–508.

(47) Sheehan, J.; Aden, A.; Paustian, K.; Killian, K.; Brenner, J.; Walsh,M.; Nelson, R. Energy and environmental aspects of using cornstover for fuel ethanol. J. Ind. Ecol. 2003, 7 (3), 117–146.

(48) Wang, M.; Mu, M.; Huo, H. Life-cycle energy and greenhousegas emission impacts of different corn ethanol plant types.Environ. Res. Lett. 2007, 2, doi:10.1088/1748-9326/2/2/024001.

(49) Houghton, R. A.; Unruh, J. D.; Lefebvre, P. A. Current land coverin the tropics and its potential for sequestering carbon. GlobalBiogeochem. Cycl. 1993, 7 (2), 305–320.

(50) Marland, G.; Schlamadinger, B. Forests for carbon sequestrationor fossil fuel substitution? A sensitivity analysis. BiomassBioenergy 1997, 13 (6), 389–397.

ES800052W

D 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. xxx, NO. xx, XXXX